Clinical Usefulness of Platelet Count

in Management of Liver Disease

Edoardo G. Giannini, MD, PhD

Cattedra di GastroenterologiaDipartimento di Medicina InternaUniversità di GenovaGenova

44th AISF Annual Meeting. Rome, February 24–25, 2011

AgendaThe role of platelets in liver disease

1. Pathophysiology of alteration of platelet count in chronic

liver disease

2. Diagnostic and prognostic usefulness of platelet count in

patients with chronic liver disease

3. Platelet count as an epidemiological marker

4. Conclusions

HaemostasisHaemostasisDiagnostic parameterDiagnostic parameter

Prognostic indicatorPrognostic indicator

Epidemiological markerEpidemiological marker

Liver damageLiver damage

Liver regenerationLiver regeneration

Platelets and Liver Disease

Pathophysiology of Alteration of

Platelet Count in Chronic Liver Disease

Pathophysiological Basis

Giannini EG, et al. Current Opin Hematol 2008; 15: 473–480

Platelets

Chronic Hepatitis

Liver Cirrhosis

Healthy Liver Portal pressure Liver function

Qamar AA, et al. Hepatology 2008; 47: 153–159

Correlation Between HVPG and Platelet Count in Cirrhosis213 patients with compensated cirrhosis and PHT, without varices

rs = -0.44; P <.0001

Not influenced by aetiology of cirrhosis

Consistent up to 5-year follow-up

Mechanism of TPO Feedback

TPO serum levels depend upon hepatic synthesis and binding toplatelet and megakaryocyte receptors

TPO

Bone marrowCirculating platelets

Healthyliver

Kaushansky K. N Engl J Med 1998; 339: 746–754

5848

3627

0

20

40

60

80

0-1 2 3 4

Liver fibrosis score

TPO

ser

um le

vels

(pg/

mL) P = .0001

P = .0001

TPO levels and liver fibrosis in patients without splenomegaly

Adinolfi LE, et al. Br J Haematol 2001; 113: 590–595

Correlation Between Serum TPO Levels and Fibrosis Stage

TPO Serum Levels Decrease as Liver Function Worsens

Giannini E, et al. Am J Gastroenterol 2003; 98: 2516–2520

R=0.489, P =.002

300

250

200

150

100

50

TPO

ser

um le

vels

(pg/

mL)

0 2 4 6 8 10 12 14 16 18 20

Aminopyrine breath test (% dose/hr at 30 min)

Thrombocytopenia

Antibody production Portal hypertension

Decreased TPO production Bone marrow suppression

Giannini EG. Aliment Pharmacol Ther 2006; 23: 1055–1065

Pathophysiological Basis

Diagnostic Use of Platelet Count in

Chronic Liver Disease

Diagnostic Use of Platelet Count in CLDAdvanced fibrosis

Cirrhosis

Response to antiviral therapy

Presence and degree of PHT

Platelets

Cirrhotic (F3-F4), n=100

Noncirrhotic (F1-F2), n=91

Likelihood Ratio: 0.91 11.53 8.45 2.12 1.30 1.10

1

38

65

8492 96

1 3 8

40

71

90

0

20

40

60

80

100

< 40,000 < 100,000 < 160,000 < 220,000 < 280,000 < 340,000Platelet count threshold (cells/µL)

Patie

nts

(%)

Correlation Between Liver Disease Severity and Platelet Counts

Bashour FN, et al. Am J Gastroenterol 2000; 95: 2936–2939

Platelet Count and Severity of Fibrosis in Chronic Hepatitis C

Giannini EG, et al. J Clin Gastroenterol 2006; 40: 521–527

Ishak fibrosis score

450000

400000

350000

300000

250000

200000

150000

100000

50000

0

Pla

tele

t cou

nt (n

/mm

3)

0 1 2 3 4 5 6

Italian series (n=309)US series (n=90)

METAVIR Fibrosis score

450000

400000

350000

300000

250000

200000

150000

100000

50000

Pla

tele

t cou

nt (n

/mm

3)

0 1 2 3 4

rs = -0.377; P <.001 rs = -0.498; P <.001

Platelet count and significant fibrosis: transportability of results

Diagnostic Accuracy of Platelet Count in Chronic Hepatitis C

Giannini EG, et al. J Clin Gastroenterol 2006; 40: 521–527

AUC = 0.733

+LR = 3.21

-LR = 0.47

AUC = 0.900

+LR = 6.17

-LR = 0.21

Cut-off = 163,000/mm3 Cut-off = 141,000/mm3

Italian and US cohorts = 409 patients

Cirrhosis = 19.1%Significant fibrosis = 42.8%

Martinez SM, et al. Hepatology 2011; 53: 325–335

Performance of Non-invasive Methods to Assess Fibrosis in Chronic Viral HepatitisScore Serum markers Etiology n ≥F2 (%) AUC≥F2 F4 (%) AUCF4

Fibrotest GGT, haptoglobin, bilirubin, apoA1, alpha2-macroglobulin

HCV 2,342 33-74 0.74-0.89 3-25 0.82-0.92

HBV 218 56-68 0.77-0.85 15-20 0.76-0.87

Forns Age, GGT, cholesterol, platelets

HCV 1,982 32-59 0.75-0.91 3-20 -HBV 456 56-74 0.63-0.72 15-26 0.81

APRI AST, platelets HCV 3,160 27-74 0.69-0.88 3-25 0.61-0.94HBV 886 56-79 0.72 15-20 0.64-0.76

FIB-4 Age, ALT, AST, platelets HCV/HIV 1,778 21-36 0.74-0.85 7 0.91HBV 776 65-79 0.74-0.86 15-34 0.80-0.93

Fibrometer Platelets, PT, macroglobulin, AST, hyaluronate, age, urea

HCV 1,039 41-56 0.78-0.89 4-15 0.94HBV 108 56 0.74 15 0.89

Hepascore Age, sex, alpha2-macroglobulin, hyaluronate, bilirubin, GGT

HCV 1,660 39-79 0.74-0.86 6-34 0.80-0.94

ELF NT-PIIIP, hyaluronate, TIMP-1, age

HCV/HBV 1,346 27-64 0.77-0.87 12-16 0.87-0.90

Thrombocytopenia and Degree of Portal Hypertension

Qamar AA, et al. Clin Gastroenterol Hepatol 2009; 7: 689–695

213 patients with compensated cirrhosis and PHT, without varices

0

5

10

15

20

25

Normal

platelet count

Thrombocytopenia

HV

PG

(mm

Hg)

% o

fpat

ient

sw

ithH

VP

G >

10m

mH

g

0

20

40

60

80

100

<150,000/mm3

Platelet count>150,000/mm3

32%

61%

P = .003

Platelet Count for the Noninvasive Diagnosis of Oesophageal Varices

Platelet Count for the Noninvasive Diagnosis of Oesophageal Varices

Qamar AA, et al. Clin Gastroenterol Hepatol 2009; 7: 689–695

AUC = 0.630(95% CI, 0.554 – 0.706)

Median platelet count at time of varices development = 91,000/mm3

(IQR = 65,000 – 123,000/mm3)

0

5

10

15

20

25

30

35

40

45

<80 80-100 100-130 130-150 150-200 >200

Variceal haemorrhage

Large varices

Small varices

Platelets in thousands

Num

bero

f pa t

ien t

s

15%

0

50

100

150

200

250

Day 0Week 4Week 8Week 12End of treatmentWeek 4 posttreatmentWeek 24 posttreatmentIFN alfa-2b IFN alfa-2b

+ RBVPegIFN alfa-2a

+ RBVPegIFN alfa-2b

+ RBV

* * *

* * *

*

* * * ** * *

*

Med

ian

Plat

elet

s (x

103

cells

/µL) *P < .05 vs baseline

Schmid M, et al. Gut 2005; 54: 1014–1020

Platelet Count During Antiviral Therapy in HCVHaematopoiesis evaluated in 4 treatment groups from 6 prospective trials (n = 133)

Platelet Count and Null Response to Antiviral Therapy

Lindsay KL, et al. Clin Gastroenterol Hepatol 2008; 6: 234–241

Variable Odds ratio (95% CI) P value

Age (per 5 y) 0.95 (0.92-0.98) .005

African American 2.75 (1.55-4.87) .0005

Previous combination therapy 1.78 (1.04-3.07) .04

Genotype 1 5.19 (1.97-13.67) .0009

Albumin (per 0.5 g/dL) 0.35 (0.19-0.63) .0005

Weight loss to week 20 (per 3 kg) 0.90 (0.84-0.95) .0003

ALT (xULN) decrease to week 20 0.83 (0.72-0.96) .01

White blood cell decrease to week 20 (per 1,000/mm3) 0.76 (0.64-0.90) .002

Platelet level decrease to week 20 (per 40,000/mm3) 0.99 (0.98-0.995) .0002

Multivariate analysis of baseline and week-20 factors associated withnull versus full or partial response (HALT-C Trial)

Prognostic Use of Platelet Count in

Chronic Liver Disease

Prognostic Use of Platelet Count in CLDPlatelets

•Variceal bleeding

• SBP

• Non-hepatic surgery

• Liver transplantation

• Compensated disease

• Advanced disease

Cirrhosis complications Surgery Survival

Thrombocytopenia and Platelet Count Cut-off

Pre

vale

nce

(%)

<150 x 109/l <50 x 109/l <20 x 109/l

Liver cirrhosis

9

61

111

83

3

19

90

Giannini E, et al. Gut 2003; 52: 1200–1205

Prognostic Indicators in Cirrhosis of the Liver

D’Amico G, et al. J Hepatol 2006; 44: 217–231

Variable Good studies in whichvariable was amongfirst 5 significant ones(n)

Good studiesevaluating variable(n)

Good studies in which variablewas among first 5/total of studies(%)

Child-Pugh score/class 13 20 65

Bilirubin* 11 23 48Albumin* 11 23 48Age 11 28 39Prothrombin time* 8 21 38Encephalopathy* 7 14 50Ascites* 4 14 29Gender 5 23 22BUN 3 4 75Platelets 3 10 30

Variables that were most commonly found to be significant predictorsof death assessed in 31 ‘good’ studies

Liangpunsakul S, et al. Am J Med Sci 2003; 326: 111–116

Hypersplenism and Incidence of Oesophageal Varices Bleeding

Severe hypersplenism defined as platelet count <75,000/mm3 and/or WBC count <2,000/mm3

Incidence of variceal bleeding:

19% versus 5% (P = .0001)

HR = 4.1 (1.7-10.0, 95%CI), P = .002

Liangpunsakul S, et al. Am J Med Sci 2003; 326: 111–116

Hypersplenism and Incidence of Spontaneous Bacterial Peritonitis

Severe hypersplenism defined as platelet count <75,000/mm3 and/or WBC count <2,000/mm3

Incidence of SPB:

16% versus 3% (P = .003)

HR = 8.0 (3.1-20.5, 95%CI), P = .0001

Fattovich G, et al. Am J Gastroenterol 2002; 97: 2886–2895

Prognostic Meaning of Platelet Count in Compensated Viral Cirrhosis

Outcome Variable Regression coefficient SE P value

HCC Age 0.061 0.017 0.000Bilirubin 1.899 0.715 0.008Albumin -0.064 0.030 0.034Viral status 0.425 0.324 NS

Decompensation Platelets -0.005 0.002 0.024Albumin -0.083 0.023 0.000Gammaglobulin 0.043 0.018 0.018AST/ALT ratio 1.525 0.486 0.002Viral status -0.523 0.234 0.026

Survival Age 0.061 0.014 0.000Sex 0.627 0.303 0.038Platelets -0.006 0.003 0.018Albumin -0.110 0.024 0.000Viral status 0.367 0.271 NS

Child-Pugh class A (n=297). Follow-up 79 months (6 – 191 months)

Thrombocytopenia and Prognosis of Compensated Cirrhosis

Qamar AA, et al. Clin Gastroenterol Hepatol 2009; 7: 689–695

Risk factors for death or OLT Hazard ratio (95% CI) P valueThrombocytopenia at baseline 4.4 (1.3 – 13.6) .0191

Leukopenia at baseline 1.8 (1.0 – 3.3) .083Baseline HVPG 1.1 (1.0 – 1.2) .0132Child-Pugh score at baseline 1.6 (1.1 – 2.2) .0105

P =.034

Normal platelet count

Thrombocytopenia

Years

Surv

ival

Liangpunsakul S, et al. Am J Med Sci 2003; 326: 111–116

Hypersplenism and Overall Survival in Advanced Liver Disease

Severe hypersplenism defined as platelet count <75,000/mm3 and/or WBC count <2,000/mm3

HR = 2.0 (1.2-3.4, 95%CI), P = .008

Median survival:

32 versus 47 months (P = .03)

Perkins L, et al. Clin Gastroenterol Hepatol 2004; 2: 1123–1128

Prognostic Meaning of Platelet Count in Cirrhotic Patients Undergoing Surgery

Variable Odds ratio P value

International normalized ratio >1.2 16.9 <.001

Bilirubin level > 1.0 mg/dL 7.1 .01

Creatinine level > 1.4 mg/dL 11.8 .04

Platelet count < 150 × 103/mm3 8.3 .04

ALT level > 40 or AST level > 30 U/L 4.5 NS

Albumin level < 3.4 g/dL 3.9 NS

Hematocrit level < 37% 2.2 NS

White blood cell count > 10.8 × 103/mm3 1.7 NS

Alkaline phosphatase > 105 U/L 1.2 NS

Hemoglobin < 12 g/dL 1.2 NS

Patients undergoing cholecystectomy: predictors of 90-day morbidity

Platelets and Liver Transplantation

Pereboom ITA, et al. Liver Transpl 2008; 14: 923–931

Platelets Count Dynamics During and After Liver Transplantation

Nascimbene A, et al. J Hepatol 2007; 47: 651–657

FFP Transfusion-related Graft-related

Platelet activation & consumption

Prognostic Meaning of Platelet Count After Liver Transplantation

Pereboom ITA, et al. Liver Transpl 2008; 14: 923–931

90-days survival, n = 449

P = .01

•Platelet consumption/chronic DIC

•Poor liver function/low TPO

•Small-for-size/portal hypertension

Platelet Count

as an

Epidemiological Marker

Wang CS, et al. Clin Infect Dis 2004; 39: 790–796

Risk Factors for Thrombocytopenia in Community Surveys

Risk factor Odds ratio 95 % CI

Anti-HCV positivity 6.0 (3.2 – 11.2)

Age ≥65 years 4.3 (2.0 – 9.5)

ALT level ≥ 40 or AST level > 30 U/L 2.1 (1.1 – 3.9)

HBsAg positivity 0.9 (0.4 – 2.1)

Analysis of risk factors for thrombocytopenia (<100,000/mm3)

in a community of hyperendemic HBV and HCV infection

1,690 subjects. Prevalence of HCV = 17.4%. Prevalence of HBV = 9.2%

National Health and Nutrition Examination Survey III (n = 16,196)

Prevalence of Anti-HCV Positivity by Platelet Count

0

10

20

30

40

50

60

<50 50-100

100-149

150-190

200-240

250-290

300-340

350-390

400-440

450-500

>500

Platelet count (x109/L)

Prop

ortio

nof

anti-

HC

Vpo

sitiv

e (%

)

Streiff MB, et al. Hepatology 2002; 35: 947–952

Overall HCV prevalence = 2.4%

Significantly greater prevalence

Huang YC, et al. J Gastroenterol Hepatol 2011; 26: 129–134

Platelet Count as a Tool for Hepatocellular Carcinoma Screening

High anti-HCV prevalence, elderly residents, low resources (n = 1,002)

Screening strategies

Lu SN, et al. Cancer 2006; 107: 2212–2222

HBsAg, anti-HCV, AFP (>20ng/mL), AST (>40U/L), ALT (>40U/L), and

platelet count (<150,000/mm3)

16 HCC cases

Any positive → Liver US

(n = 527)

AFP (>20ng/mL) and

platelet count (<150,000/mm3)

14/16 HCC cases (88%)

Any positive → Liver US

(n = 215)

Cost = 44,816 USD Cost = 18,044 USD (- 60%)

Summary and Conclusions

Clinical Usefulness of Platelet Count in Chronic Liver Diseases

Increase in portal pressure and decrease in liver function are the main pathophysiological mechanisms responsible for decreased platelet counts in CLD patients

In these patients, platelet count can be used as a diagnostic parameter, a prognostic indicator, and an epidemiological marker

From the clinical standpoint, platelet count – alone or in composite scores – represents an aid to identify advanced fibrosis and cirrhosis, to foresee the occurrence of complications of cirrhosis and establish patients prognosis, and may help targeting screening programs for hepatocellular carcinoma

PLATELETS

AGGREGATE

The Hepatologist’s View of Coagulation